Inferring gene regulatory networks using differential evolution with local search heuristics

Area of Science:

• Systems Biology
• Computational Biology
• Bioinformatics

Background:

• Inferring gene regulatory networks is crucial for understanding cellular mechanisms.
• Traditional methods often struggle with accuracy and parameter estimation.

Purpose of the Study:

• To develop an advanced memetic algorithm for gene network structure evolution and parameter inference.
• To enhance gene network model selection using Information Criteria over Mean Squared Error.

Main Methods:

• A memetic algorithm incorporating hill-climbing local search was employed.
• Decoupled S-system formalism was used for biomolecular interaction modeling.
• Information Criteria replaced Mean Squared Error for fitness evaluation.

Main Results:

• The algorithm accurately inferred gene network topology and regulatory parameters across various experimental conditions.
• Performance was sensitive to data quantity and noise levels.
• The Information Criteria-based fitness function outperformed conventional methods in accuracy.

Conclusions:

• The proposed memetic algorithm offers a robust approach for gene network reconstruction.
• This method enhances the identification of network topology and parameter estimation.
• Applied to yeast cell-cycle data, it successfully reconstructed key regulatory networks.